1Mby1M Virtual Accelerator AI Investor Forum: With Ashish Gupta, Partner at Clearvision Ventures (Part 1)

Ashish Gupta, Partner at Clearvision Ventures, has been in the industry for a long time and has an interesting perspective on AI investing.

Sramana Mitra: So we’re going to start today’s session with a conversation with Ashish Gupta, partner of Clear Vision Ventures. Ashish has had a long career in venture capital and I’ve known him for a very long time as a friend.

We talk periodically and exchange notes. So this is a wonderful opportunity to share one of our catching up conversations with you.

So welcome Ashish, great to see you.

Ashish Gupta: Thank you for having me.

Sramana Mitra: So Ashish, of course the world is hyperventilating on AI. So let’s start with that discussion. What do you see? How do you parse? What is your AI investment thesis?

Ashish Gupta: So I think even AI, and then more recently, Gen AI, and as we’ll go further, I’m sure you’ll parse those two out. I’m a database guy by training. I’m a big believer in the “blind men and the elephant” story paradigm, where five blind people each feel different body parts of an elephant and conclude that the elephant is a totally different object.

I think I am one such blind person who touches and feels a whole bunch of things. My perspective is strongly shaped by the database bias I carry from my past, which means I tend to view technologies through that particular lens.

AI is particularly interesting in that regard, because it has replaced a lot of tasks that people like me used to do – data mining and data processing – with brute force compute. If you could comprehend three dimensions, we will instead figure out thirty dimensions at one point in time by finding patterns using machine learning algorithms, which is enormously powerful. Companies that have the data can leverage that to improve their own executions at a fantastic pace. So I think it furthers the power of the already powerful in ways that are very hard to comprehend.

You can see evidence of this in the market caps of a lot of the larger companies that have taken one of the assets that they had. People often talk of data as “oil”, but I think it is a little dangerous analogy. You also need to own the refinery. Otherwise, your data is more like sludge than oil.

I sometimes wonder if the speed at which AI is advancing might actually hurt startup activity. My concern is that while AI disproportionately benefits larger companies, it could make it harder for AI-based startups to maintain a vibrant value proposition.

Sramana Mitra: Very interesting perspective.

So let’s double click down on that. We’ve had seven or eight conversations with investors so far in this AI series that we’re doing. Almost universally, everybody’s excited about vertical AI and they’re investing in vertically AI startups – people who are coming up with a small, well-defined problem and really solving that with AI in a very meaningful way.

So is your perspective significantly different from that vertically AI?

Ashish Gupta: So I think there are many ways to approach this. If we assume, for a moment, that the reason to contain the scope of AI applications is to reduce the probabilistic nature of the results while increasing the depth of the models, both of which therefore lead to better outcomes, deeper and more precise.

I actually agree with that thesis.

One of the things that has personally excited me—and this is just one perspective, following the “blind men and the elephant” analogy—is the involvement of human beings alongside AI. Companies that implement human-in-the-loop systems are particularly interesting to me because they help address some of the imprecision that AI can introduce. These systems also allow for constant refinement, which is crucial, especially for companies that lack the internal talent pool needed for that constant refinement.

When I apply this thinking to your point about vertical AI, it actually begins to add up in my head. Often, these vertical players don’t have deep technical teams. While Amazon can leverage AI effectively, if we look at a company like Pfizer—and I’m just using Pfizer as an example, with no intent to criticize their technical capabilities—I imagine they might struggle to hire the same level of AI talent as an Amazon or Facebook.

By focusing on vertical AI and adding a services component, you’re not only making solutions more precise, but you’re also helping bridge the gap between recognizing the potential of powerful technology and actually converting it into practical solutions. This is the direction I’ve been leaning toward.

This segment is part 1 in the series : 1Mby1M Virtual Accelerator AI Investor Forum: With Ashish Gupta, Partner at Clearvision Ventures
1 2 3 4